References:
Use the following command to load this dataset in TFDS:
ds = tfds.load('huggingface:empathetic_dialogues')
- Description:
PyTorch original implementation of Towards Empathetic Open-domain Conversation Models: a New Benchmark and Dataset
- License: No known license
- Version: 0.1.0
- Splits:
| Split | Examples |
|---|---|
'test' |
10943 |
'train' |
76673 |
'validation' |
12030 |
- Features:
{
"conv_id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"utterance_idx": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"prompt": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"speaker_idx": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"utterance": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"selfeval": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"tags": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}